AI Literacy Is Already Required. What Does That Mean in Practice?

The EU AI Act is often discussed as if its obligations are still somewhere in the future. For many parts of the Act, the main operational deadlines are still rolling out. But one requirement is already in force: AI literacy.

Since 2 February 2025, Article 4 has required providers and deployers of AI systems to take measures to ensure a sufficient level of AI literacy among staff and other people using AI systems on their behalf. The European Commission describes AI literacy as the skills, knowledge, and understanding needed to make informed use of AI systems and to understand the opportunities, risks, and potential harms they may create.

That sounds like training, and training is part of it. But AI literacy should not be treated as a one-off course that employees click through once a year.

AI literacy is not just knowing how to prompt

Many organisations think of AI literacy as teaching employees how to use tools better: how to write prompts, summarise documents, draft faster, and use AI for productivity. Those skills are useful, but they are only part of the picture.

The more important question is whether people understand where AI can fail. Can they recognise when an answer is too confident? Can they tell when a source does not support the claim? Do they know that a fluent summary can still miss an important caveat? Do they understand when AI output should be escalated, reviewed, or rejected? A person who can generate AI output quickly is not necessarily AI literate in a professional setting. Speed of generation is not the same as judgment about whether the output should be used, and in client-facing work the second skill matters more than the first.

Evaluation is the real skill

In client-facing work, the key skill is not only using AI. It is knowing how to question it. That means checking whether the answer is supported by evidence, whether the source material is current, whether the model has ignored contradictions, and whether the final output is appropriate for the client, context, and risk level.

AI literacy also depends on role and context. The Commission’s guidance says organisations should consider the technical knowledge, experience, education, and training of staff, as well as the context in which AI systems are used and the people affected by them.

That matters because not every team needs the same level of AI literacy. A marketing team using AI for early brainstorming does not need the same controls as a team using AI to prepare client-facing analysis, investment research, compliance materials, or public statements.

AI literacy cannot live only in a training slide

AI literacy has to appear in how work is reviewed. If a team knows AI can hallucinate but has no process for checking unsupported claims, the literacy is incomplete. If people understand that sources can be outdated but no one checks the source base before analysis, the risk remains. If everyone agrees that human oversight matters but there is no record of who reviewed the output, the process is still weak.

Practical AI literacy means teams know what to check and have a workflow that helps them check it. That includes source review, risk flags, model comparison where appropriate, and reviewer sign-off before important work leaves the organisation.

From awareness to accountability

The purpose of AI literacy is not to turn every employee into an AI engineer. It is to make sure people using AI understand enough to use it responsibly in their role.

For professional teams, that means moving from awareness to accountability. People should know when AI output is safe to use, when it needs review, and when it should not be relied on.

Qonera is built around that practical layer. It helps teams verify sources, compare model outputs, flag unsupported claims, and record reviewer sign-off before AI-assisted work reaches a client, partner, regulator, or decision-maker. For more on how the EU AI Act maps to operational workflow, see Qonera’s EU AI Act page. AI literacy is already required, and the next question is whether it is visible in the way your team actually works.

This article is for general information only and does not provide legal advice. Organisations should consult qualified legal counsel about how Article 4 and the EU AI Act apply to their specific systems, workflows, and obligations.